Oxygen and carbon dioxide targets in pediatric patients with return of spontaneous circulation after cardiac arrest (PLS): Systematic Review

Conflict of Interest Declaration

The ILCOR Continuous Evidence Evaluation process is guided by a rigorous ILCOR Conflict of Interest policy. The following Task Force members and other authors were recused from the discussion as they declared a conflict of interest:

The following Task Force members and other authors declared an intellectual conflict of interest and this was acknowledged and managed by the Task Force Chairs and Conflict of Interest committees:

CoSTR Citation

Berg KM, Holmberg M, Nicholson T, Nolan J, Reynolds J, Schexnayder S, Nation K, Soar J, on behalf of the International Liaison Committee on Resuscitation Advanced Life Support and Paediatric Task Forces. Oxygenation and Ventilation Targets in Adults and Children with Return of Spontaneous Circulation after Cardiac Arrest, Consensus on Science with Treatment Recommendations; 3 January 2020 Available from: http://ilcor.org

Methodological Preamble and Link to Published Systematic Review

The continuous evidence evaluation process for the production of Consensus on Science with Treatment Recommendations (CoSTR) started with a systematic review of oxygen and carbon dioxide targets in adults and children with return of spontaneous circulation (ROSC) after cardiac arrest (Holmberg et al, 2020, PROSPERO registration number pending, registered on September 24, 2019) conducted by the designated Expert Systematic Reviewer and Systematic Reviewer mentee, with involvement of clinical content experts from the Advanced Life Support and Paediatric task forces. Evidence for adult and paediatric literature was sought and considered by the Advanced Life Support Task Force and the Paediatric Task Force groups respectively. All data found were taken into account when formulating the Treatment Recommendations. This CoSTR focuses on the evidence at it applies to paediatric patients.

Systematic Review

Webmaster to insert the Systematic Review citation and link to PubMed using this format when it is available.Holmberg et al, Oxygen and Carbon Dioxide Targets after Cardiac Arrest: A Systematic Review and Meta-Analysis.

PICOST

The PICOST (Population, Intervention, Comparator, Outcome, Study Designs and Timeframe)

Population: Unresponsive children with sustained return of spontaneous circulation (ROSC) after cardiac arrest in any setting.

Intervention: A ventilation strategy targeting specific SpO₂, PaO₂, and/or PaCO₂ targets.

Comparators: Treatment without specific targets or with an alternate target to the intervention.

Outcomes: Clinical outcome including survival/survival with a favorable neurological outcome at hospital discharge/30 days, and survival/survival with a favorable neurological outcome after hospital discharge/30 days (e.g. 90 days, 180 days, 1 year).

Study Designs: Randomized trials, non-randomized controlled trials, and observational studies (cohort studies and case-control studies) with a control group (i.e. patients treated with no specific SpO₂, PaO₂, and/or PaCO₂ targets or an alternative target to the intervention) will be included. Animal studies, ecological studies, case series, case reports, reviews, abstracts, editorials, comments, and letters to the editor will not be included. There were no limitations on publication period or study language, as long as there is an English abstract. The population includes patients suffering from IHCA or OHCA of any origin. Unpublished studies (e.g., conference abstracts, trial protocols) were excluded. The cited systematic review was done without age restriction, and the evidence from adult studies (generally defined as >16 or ³18) is included here.

Timeframe: All years and all languages were included. Literature search updated to August 22, 2019.

PROSPERO Registration (registered Sept 24, 2019. Final registration number pending)

NOTE FOR RISK OF BIAS: In most cases bias was assessed per comparison rather than per outcome, since there were no meaningful differences in bias across outcomes. In cases where differences in risk of bias existed between outcomes this was noted. In studies that looked at both survival and survival with favorable neurologic outcome and outcome assessors were not blinded, for example, risk of bias was assessed separately for each outcome as good neurologic outcome is more susceptible to bias.

Consensus on Science

Oxygen targets:

We identified no paediatric randomized trials on this topic, and two observational studies since the prior review {Van Zellem 2015, Lopez-Herce 2014}. One of these {Lopez-Herce 2014}was deemed at critical risk of bias due to the lack of adjustment for cardiac arrest characteristics, and thus interpretation of these results is severely limited. Within these limitations, this study included 253 patients and found no association between hyperoxemia and clinical outcomes in adjusted analyses (numeric adjusted results not reported). Of all studies identified, only three pediatric studies {Del Castillo 2012 1456, Bennet 2013 1534, Van Zellem 2015 150}, including a total of 618 patients, were deemed to have only serious risk of bias, and within these only adjusted results were considered.

For the critical outcome of survival to hospital discharge with good neurologic outcome, we identified one observational study of 153 pediatric patients with ROSC after cardiac arrest in any setting {Bennett 2013 1534} providing very low-certainty evidence (downgraded for indirectness, imprecision, and risk of bias) comparing hyperoxemia to no hyperoxemia and finding no benefit (OR 1.02 [95% CI 0.46 to 2.27], 5 more per 1,000 [95% CI from 170 fewer to 202 more]).

For the critical outcome of survival to hospital discharge, we identified one observational study of 164 pediatric patients with ROSC after IHCA {del Castillo 2012 1456} providing very low-certainty evidence (downgraded for indirectness, imprecision, and very serious risk of bias) comparing hyperoxemia and normoxemia and finding no benefit, although numeric results of adjusted analyses were not reported. We identified a second study of 200 pediatric patients with ROSC after cardiac arrest in any setting {Van Zellem 2015 150} providing very-low certainty evidence (downgraded for indirectness, imprecision, and serious risk of bias) showing no association between post-ROSC pO2>200mmHg and outcome (OR 0.81; 95% CI, 0.41–1.59, absolute risk difference not calculable as survival in normoxemia group not reported).

One larger registry-based study {Ferguson 2012 335} did find that hyperoxemia was associated with higher mortality when compared to normoxemia, but although this study was much larger than the others, it was deemed at critical risk of bias. Factors that increased the concern about bias were the lack of adjustment for cardiac arrest characteristics (increasing the risk of confounding) and the exclusion of 31% of all eligible patients due to lack of an arterial blood gas within one hour of ROSC, (increasing selection bias as patients without an arterial blood gas are likely disproportionally normoxemic or hyperoxemic). Furthermore, the use of pO₂^-1 and pO₂^-2 as exposure variables, rather than pO₂ itself, is unique to this study, making it difficult to compare with other studies.

Carbon dioxide targets:

We identified no paediatric randomized trials on this topic. Two observational studies were identified {Del Castillo 2012, Lopez-Herce 2014}, one of which was published since the prior CoSTR{Lopez-Herce 2014}. Only adjusted results from these studies were considered. One study {Del Castillo 2012} including 223 patients provided very low certainty evidence (downgraded for risk of bias and indirectness) that both hypocapnia after ROSC (OR 2.71 [95% CI 1.04–7.05], 242 more per 1,000 [95% CI from 9 more to 446 more])) and hypercapnia after ROSC (OR: 3.27 [95% CI 1.62–6.61], 286 more per 1,000 [95% CI from 114 more to 423 more]) were associated with hospital mortality. The one study published since the prior review {Lopez-Herce 2014} was deemed at critical risk of bias due to lack of adjustment for cardiac arrest characteristics. Within these limitations, this study included 253 patients and found that both hypocapnia (OR 2.62 [95% CI 1.08-6.4], 233 more per 1,000 [95% CI from 17 more to 429 more]) and hypercapnia (OR 2.0 [95% CI 1.01-3.97], 166 more per 1,000 [95%CI from 2 more to 332 more]) one hour after ROSC were associated with hospital mortality, when compared to normocapnia after ROSC. The available evidence on the effect of hypercapnia or hypocapnia in adults is inconsistent, with the randomized trials done to-date showing no effect.

Treatment Recommendations

We suggest that rescuers measure PaO2 after ROSC and target a value appropriate to the specific patient condition. In the absence of specific patient data, we suggest rescuers target normoxemia after ROSC (weak recommendation, very-low-quality evidence). Given the availability of continuous pulse oximetry, targeting an oxygen saturation of 94-99% may be a reasonable alternative to measuring PaO2 and titrating oxygen when feasible to achieve normoxia.

We suggest that rescuers measure PaCO2 after ROSC and target normocapnia (weak recommendation, very-low-certainty evidence). Consider adjustments to the target paCO2 for specific patient populations where normocapnia may not desirable (e.g. chronic lung disease with chronic hypercapnia, congenital heart disease with single ventricle physiology, increased intracranial pressure with impending herniation)

Justification and Evidence to Decision Framework Highlights

Narrative Reporting of the Evidence to Decision Framework Incorporating Values and Preferences and other domains included in the framework, by Task Force Chairs. Technical Remarks refers to details that helps to provide specificity for the recommendation based on the current science i.e. dosing or timing.

Oxygen targets:

Accurate targeting of post-ROSC normoxemia might be achievable and acceptable guided by pulse oximetry in the in-hospital setting, but its use in the prehospital setting has not been studied and is not without risk of inadvertent patient hypoxemia. Any titration of oxygen delivery to children after ROSC must be balanced against the risk of inadvertent hypoxemia stemming from overzealous weaning of FiO2 given the known risks of hypoxemia and uncertain risks of hyperoxia. Further challenges include identifying what the appropriate targets should be for specific pediatric patient subpopulations (e.g., infants and children with cyanotic heart disease)

Carbon dioxide targets:

Accurate targeting of post-ROSC normocapnia might be achievable and acceptable in the in-hospital intensive care setting. Serial assessment of ventilation is facilitated by arterial catherization, which may also be beneficial for targeting post-ROSC blood pressure targets. Correlation of PaCO2 and ETCO2 may allow ongoing monitoring of ventilation when continuous capnography is available. Further challenges include identifying any modified PaCO2 targets for specific pediatric patient subpopulations (e.g., infants and children with suspected increased intracranial pressure).

Knowledge Gaps

Knowledge Gaps Template for Task Force chairs

1. There are no paediatric randomized trials comparing either oxygen or carbon dioxide strategies.
2. How PaCO₂ targets should be adjusted in those with chronic CO₂ retention is unknown.
3. Whether adjusting arterial blood gas analysis to 37 °C or to a patient’s current temperature is preferred is unknown.

Attachments

PLS-Et D 544 845 O2 PEDS

PLS-Et D 544 845 CO2 PEDS

References

References listed alphabetically by first author last name in this citation format

1. Bennett, Kimberly Statler; Clark, Amy E.; Meert, Kathleen L.; Topjian, Alexis A.; Schleien, Charles L.; Shaffner, Donald H.; Dean, J. Michael; Moler, Frank W.; Pediatric Emergency Care Medicine Applied Research, Network. Early oxygenation and ventilation measurements after pediatric cardiac arrest: lack of association with outcome. Critical care medicine 2013;41(6):1534-42 2013

2. Del Castillo, Jimena; Lopez-Herce, Jesus; Matamoros, Martha; Canadas, Sonia; Rodriguez-Calvo, Ana; Cechetti, Corrado; Rodriguez-Nunez, Antonio; Alvarez, Angel Carrillo; Iberoamerican Pediatric Cardiac Arrest Study Network. Hyperoxia, hypocapnia and hypercapnia as outcome factors after cardiac arrest in children. Resuscitation 2012;83(12):1456-61 2012

3. Ferguson, Lee P.; Durward, Andrew; Tibby, Shane M. Relationship between arterial partial oxygen pressure after resuscitation from cardiac arrest and mortality in children. Circulation 2012;126(3):335-42 2012

4. Lopez-Herce, Jesus; del Castillo, Jimena; Matamoros, Martha; Canadas, Sonia; Rodriguez-Calvo, Ana; Cecchetti, Corrado; Rodriguez-Nunez, Antonio; Carrillo, Angel; Iberoamerican Pediatric Cardiac Arrest Study Network. Post return of spontaneous circulation factors associated with mortality in pediatric in-hospital cardiac arrest: a prospective multicenter multinational observational study. Critical care (London, England) 2014;18(6):607 2014

4. van Zellem, Lennart; de Jonge, Rogier; van Rosmalen, Joost; Reiss, Irwin; Tibboel, Dick; Buysse, Corinne. High cumulative oxygen levels are associated with improved survival of children treated with mild therapeutic hypothermia after cardiac arrest. Resuscitation 2015;90():150-7 2015